reply to post by spikedmilk
Between the JPEG and the GIF files you chose the best, the GIF, because this type of file does not suffer from compression artifacts like JPEG.
But the GIF file had its brightness and contrast changed to create a more visible image (when compared to the original data), and that change is made
to give a good look to the whole image, so some parts of the image have certainly lost some detail.
If you want to make a more in-depth analysis you should use the original IMG or IMQ file. This file has the original data, and even the processed
version has not had its colours adjusted. Using this file you can select the area you want to work with and apply the changes only to that area,
resulting in the best possible version of that area.
To see (and work with) IMG and/or IMQ files you need NASAView (available
here, but they
ask for name and e-mail address, and I think not all people want to provide it
) or The GIMP and a plug-in specifically created for that.
The plug-in does not work with the latest GIMP version, you need version 2.2.xx (available
here),
and the plug-in is available
here.
The GIMP is a very capable program, but a little strange to work with.
For the "sand castle", for example, after opening the IMG file in The GIMP I selected that area and choosing the menu "Image/Crop Image" to work
only with that smaller area. Then I chose "Tools/Color Tools/Levels" to change the brightness and contrast.
In this tool I moved the markers for the white point (1 in the image bellow) and for the black point (2) to just before reaching the area of the
histogram where there is the brighter and the darker tones, respectively.
To be easier to see where those points are I changed the histogram view to logarithmic (3) and marked the preview to see if I could make a better
image.
Moving a little the middle pointer (5, the gamma value) I finally achieved an image with which I was happy. It looked good and it had no extreme (too
white or too black) areas, so it probably hadn't lost any data.
So, starting with this image from the original IMG file I reached an optimised image for the are of the "sand castle".
I opened both images in Paint Shop Pro 6 and saw that both versions have 89 different shades of grey, so it did not loose any information, it was only
made more visible. As a comparison, the exact same area from the GIF file has only 46 shades of grey, so even that version had already lost almost
half the original data.
(original "sand castle" from IMG file)
(my optimised version)
(original from the GIF file)
Once again sorry by over-reacting to your images.
Happy analysis.
PS: I would like to point that I am no expert in image manipulation and/or analysis, I just do what I have learned gives the best results, if anyone
has any better way of doing it I would like to learn it, learning is one of the things I like the most.